About our guest:
Danielle Wallace is the Chief Learning Strategist at Beyond Role Plays, an organization that builds custom AI-powered role play training for sales and service teams, so employees can practice real conversations before having them with actual customers. The scenarios are designed by learning experts and tailored to each organization.
More about Danielle Wallace
More about Beyond Role Plays
AI + L&D Community of Practice AI + Learning & Development Community of Practice | Groups | LinkedIn
Transcript of the Interview
{Edited for Ease of Reading}
Petra Mayer
Hello and welcome everybody to today’s session on AI in learning and development. My name is Petra Mayer of Petra Mayer & Associates Consulting and today I have with me Danielle Wallace. Danielle is from Beyond the Sky Custom Learning. She is a custom learning specialist and a specialist in AI. Today’s conversation is about using AI in learning and development, understanding some of those risks and how to mitigate them, talking about some tools, tips, and best practices. Welcome to the show, Danielle. So glad to have you here.
Danielle Wallace
It’s so lovely. Thank you so much.
Petra Mayer
So Danielle, before we start off, everybody’s talking about AI and all kinds of life situations and of course it’s also really impacting the learning and development area in HR and in corporations. How is it changing the world of learning and development now that we have access to AI?
Danielle Wallace
So the tech is rapidly accelerating beyond what we can actually foresee. The tech is even faster than what the headlines are suggesting. And yet for the everyday person, and especially for the learning and development professional, we’re moving at a much slower rate than what the headlines are suggesting. So when you’re seeing things about what’s new and what’s being replaced and what’s coming, we’re not really there. That’s the interesting thing I would say. I tell people who get worried: “Don’t worry, you’re not behind, everybody’s behind.” But at the same turn, I will also strongly say you better step up, because the pace is getting quicker and we have an ability to contribute more by benefiting from some of the efficiencies with AI.
Petra Mayer
So I heard you say that in L&D we’re actually even more behind than in other areas. Why is that? What do you think is holding us back?
Danielle Wallace
So I’ve seen this as a systemic challenge throughout the decades. I saw this from a marketing standpoint with marketing technologies. Learning and development implemented use cases for the exact same tech eight years later. Eight years later. And I was really hoping it wouldn’t happen with AI, but I’m definitely seeing the lag. I think it’s for many reasons. In other industries there is more of a push towards test-and-learn, there are budgets to support that, there’s more data inherent in the nature of those roles. In marketing, for example, that’s the role. And then thirdly, the reason is that we feel like we’re people people, we actually like to interact, like the human connections. We don’t define our personas around how we’re using technology, whereas other industries do. There’s always been more integration of technology, more data, and more focus on business goals: let’s sell more, let’s ship more, let’s drive margins. That’s always been inherent in a lot of other functions which have embraced technology more. And we’re definitely seeing this if I compare marketing to learning and development, there’s a pretty big gap there. And that’s exciting, because it means as learning and development professionals, we have an ability to up our game. These are things that have already been proven, and this is our ability to shine.
Petra Mayer
It’s really interesting because as I hear you talk, I’m thinking about my clients who have a learning management system project. It’s such a big project and it tends to take a long time, and there’s a hesitation to replace your LMS. Now of course a lot of LMSs have AI incorporated or are working with AI tools. So if the LMS they have doesn’t have that, it’s a big lift for an organization to update. But where do you see the biggest opportunities for AI to add real value in L&D? Not just AI because everybody else is using it, but we want real business outcomes. How can L&D support business objectives better by using AI?
Danielle Wallace
So I see the practical use as using AI to help springboard the things we have at a task level that maybe aren’t the best use of our time. Data analysis, which we should be doing if we’re not already, working with subject matter experts to consolidate meeting outcomes, helping define task-level work. I think that’s where there actually is a big impact, and I don’t say that lightly. Because the tremendous gains, and it comes with some pitfalls, the tremendous gains are in redesigning the entire task, the entire workflow. That’s where we will get to. L&D is not ready for that yet, and that’s okay. The real benefit is redesigning the entire task flow. That’s what some of the work I’m doing with my team, in a very careful way, looking at: if this is our outcome, the path to get there might be completely different. Just breaking everything down, redesigning fully at the task level, and being very careful where we are using our human judgment to discern. And notice that in those somewhat trite examples I gave of supporting course creation, what I’m not saying is to make the bigger leaps. Because on a mass scale, using AI to create a course, I’ve seen the outputs of that without this level of discernment, and it leads to some pretty bad results. Polished fluff. Making bad training faster. So we have a lot of ability to grow in learning and development, and where we are right now, at these first steps, is the right way to drive efficiency gains without using AI to replace the critical thinking aspects, which would actually be very easy to offload.
Petra Mayer
That’s really interesting, because what I heard you say is that the big benefit will come from really reimagining the whole workflow. That is scary. And to do that at a time when L&D teams tend to be really pressured to create a lot of good learning with very little resources, redesigning while you’re doing that is quite a lot. So no wonder there’s a bit of a delay in L&D right now. And you also mentioned there is a danger that we can make bad training faster. Tell us a little bit more about that. How does that happen?
Danielle Wallace
So if you ask the average person, if every listener looks within their talent and community, you ask anybody what training is, they will tell you it’s a teacher lecturing, there’s a quiz at the end. It’s a video, there’s a quiz. The world and the internet perpetuates this idea that learning is teaching. And that’s actually part of the danger, because all the LLMs are trained on that. So even just using a regular LLM, you have to work really hard to break out of that. Otherwise all you’re going to get is lecture courses with a quiz at the end. But you amplify that further. There are a lot of brilliant tech companies that I really respect and admire. Tech companies, though, that don’t know learning and development. When they create tech, it’s likewise built around this idea that training is a course, it’s content, there’s a quiz at the end. So we have a lot of great tech companies making bad training faster. The worst example I would say is think about the ability to have a lecturer deliver a morning course. There’s so much of that out there. So many lectures delivering boring courses whether in person, virtual, or pre-recorded. Now I can use AI instead of doing that. I can create an AI avatar delivering my 27-minute lecture course with a quiz at the end. It wasn’t good to begin with. Now we have tech that makes that faster. And it’s not just the videos. Now we can create an e-learning version of a bad course, made faster. We can create polished fluff. We can make that faster. We can create training courses on things that don’t need to be trained, just because we can. The world doesn’t need more training. It needs more effective training to derive the business outcome. And unfortunately, technology is masking that because we can do a lot faster. But are we doing the right things?
Petra Mayer
So ultimately, scale is always an issue for learning and development and has been an issue. Limited people, limited resources to create training. Now we’re throwing AI at it so we can create training faster, but if the training isn’t well designed, isn’t effective, doesn’t meet the learner where they’re at, then we’re just accelerating the creation of bad stuff. And ultimately, all this bad stuff ends up in our LMSs and clogs it for people who can’t even find anything anymore. So lots of risks there. When we think about these risks, what are the common mistakes organizations make when they rush into AI because everybody else is doing it? I’ve got to do it too, or otherwise I’ll be behind? What are those common mistakes?
Danielle Wallace
So the common mistakes from a learning and development lens, thinking firstly through the lens of content creation. The world thinks training is content. Using AI to create a course. First: do you need the course? Are the performance objectives right? And then secondly, the discernment that’s needed to really pick through the outline, pick through the objectives. Is this meeting the business goals? That would be the first one. The second is a fallacy I’ve seen very smart, experienced, senior instructional designers fall into: taking disparate topics and using AI to create an outline because we’ve been told it can do some of this work, it can create learning objectives for you. But as I already mentioned, we still haven’t negated the need for upfront performance consulting. We can now create a course outline with AI. Great. But it takes an even more critical lens to parse out that outline and really think: do these topics make sense together? Is this timing realistic? Are these activities in conflict with each other? Is this even the right application activity? Why are there no application activities? There’s such a level of discernment that’s needed and it’s harder than ever, because the output is polished. If this was raw, our brains would go, okay, this is raw, I can parse through this. But it’s so polished it makes it even harder. You take that a step further and say this is not just an outline but it’s in a published course format. It’s even harder, because it’s just so pretty and beautiful. But these topics perhaps don’t even fit together, or they don’t belong in an e-learning at all. There are so many issues there, because the level of human discernment and critical thinking skills we already had as a challenge becomes even more important to exercise.
Petra Mayer
It makes me think of a really well-designed infographic where the data is actually rubbish. Because it looks so well-designed, we just believe what’s in it. And I can see that it’s much harder to be critical of something that is well-designed and polished than something that’s in a rough stage. So having said that, and you’ve alluded to business objectives a number of times, before an organization adopts any AI tools, what should they do and how should they review their current learning strategy so that they’re better equipped to utilize the right tools for the right tasks?
Danielle Wallace
So it’s really important to start with the strategy. Really think about what business outcomes are needed. How am I aligned to the business? What is needed there? And this is not new, but in a world where we’re more pressed for time, it becomes more important. Stop. What is the learning strategy? Am I just adopting tools for the sake of tools? Am I doing training for the sake of training? Which in itself is another entire issue that’s quite common. And then based on that, what skills does my organization need? It’s classic learning strategy consulting. What’s needed to get to that end outcome. That part becomes even more important, and it’s the same challenge, the same fight that we’ve had for a while. I empathize, because it’s not easy, but it’s more important than ever. And then with that in hand, when we are using AI at the individual level within the learning and development team, it’s thinking about where AI is best used and where it isn’t. For example, and especially for a more senior instructional designer, a tip I would tell people is actually don’t use AI as your starting point thought partner. That’s dreadful. You’re probably going to get an output that’s outdated. A content-driven, lecture-filled course. Don’t start with AI. Start with yourself and exercise your own critical thinking. Really keep working your own critical thinking. And then use AI to fill in the gaps. To make the repetitive parts faster. And then secondly, from the strategy level, look at what is the business need and then what are some of the applications that support that. If the goal is a specific outcome, what is the best way of achieving that? Is it e-learning, is it applying that skill in the real world and getting feedback? The importance of having the right learning strategy becomes much more important than it ever has been.
Petra Mayer
And what I can also imagine is that many senior learning and development specialists are under a lot of pressure to add AI and speed up the process. And in reality, from what you just said, it makes more sense to slow down and really rethink the strategy. Can we use the tools available today to be not only more efficient but also more effective? What are your thoughts on that?
Danielle Wallace
It’s about being intentional about where we are spending our time. Some of the things that can’t be replaced are the human connections. Aligning with your business partners, aligning with the stakeholders. I’ve seen when that’s been automated, and I’ve seen courses that are beautiful, instructionally sound, but built on the wrong need because that part was missed. So it’s being very intentional about where we’re putting our time. That remains really critical. Designing and deriving the right learning solution is important. Again, this is an exercise of human judgment. In contrast, where we could speed up is at the task level, things like changing tone within a piece, generating visuals, generating videos. These are all task execution level. Let’s automate those things that don’t require a senior instructional designer. If everybody, no matter what the role, thinks of it from that lens, that provides a nice litmus test.
Petra Mayer
There is in general a fear right now that AI will replace people, especially at the entry level. So how can we best use AI to support the L&D team? Not to replace them, but to allow people to come in at the entry level, go up through the ranks, build their experience, and still have the experience of those entry-level tasks? You said we could sometimes automate those tasks through AI, but we still need people to actually learn the process. What are your thoughts on that?
Danielle Wallace
I’m on one hand saying caution about using AI, and on the other hand I’m a strong supporter of using it, and I get the dichotomy of that. I definitely think there’s still a role for junior people within learning and development. But maybe I’d rather not have them spend their time copying and pasting from a design document into a course, or searching image banks to find the right image that we never quite find anyway, and those aren’t good placeholder images to begin with. That’s not where skill capability gets built. Skill capability comes through conversations with the business from a learning consulting lens, or from an instructional designer lens, deriving the right strategy, thinking about what application activities could be meaningful. Those are skills that need to be developed and can still be developed with our junior people, even if we’re using AI in the process of that.
Petra Mayer
So what advice would you give to smaller organizations that may not have access to as many tools? Where should they start?
Danielle Wallace
Firstly, start. Doing anything is better than doing nothing. You can engage and learn from that experience, even whether or not you decide to release whatever you’re creating. So firstly, just start. Secondly, through this lens of caution I’ve been sharing, to avoid making bad training faster, especially within smaller organizations where time is more pressed, is it really of value to create a course that isn’t actually serving the business need? Yes, you can create it fast, but is it really serving the need? I think if a senior instructional designer creates a course in seven hours and it looks beautiful but missed the upfront learning consulting piece, there’s no point. Better to have nothing, if you take it from the macro lens, than to have something ineffective. So be cautious about where you are optimizing without losing sight of the fact that it has to be effective. We can’t just have a world of lecture-filled courses with knowledge checks at the end if you’re truly trying to drive skills.
Petra Mayer
I really appreciate your views on that. You’re fully supportive of AI and you see the risks, and you’re trying to support your clients to mitigate them and be aware of the potential to make bad training faster. Do you have any final words of wisdom you’d like to share with our audience before we sign off?
Danielle Wallace
Embrace AI. These are tools we should be incorporating into our way of life because that’s where we gain added benefit. But do so realizing that your own human judgment and critical thinking become extra important. What I’m really committed to, within the learning platforms I’ve built, is driving that critical thinking component, because it’s so easy to offload. I don’t want to lose sight of that while still encouraging everybody to keep pushing on the automation of task-level work.
Petra Mayer
Thank you so much for your insights, Danielle. How can listeners find you and get in touch?
Danielle Wallace
I can be reached on LinkedIn at Danielle Wallace, or you can find me online at beyondroleplays.com.
Petra Mayer
Thank you so much. We will definitely add those links to the post.
